Research article

Least energy sign-changing solution for a fractional p-Laplacian problem with exponential critical growth

  • Received: 21 August 2022 Revised: 14 September 2022 Accepted: 21 September 2022 Published: 27 September 2022
  • MSC : 35J60, 35J92, 35R11

  • In this paper, we consider the following fractional p-Laplacian equation involving Trudinger-Moser nonlinearity:

    (Δ)sN/su+V(x)|u|Ns2u=f(u)  in RN,

    where s(0,1),2<Ns=p. The nonlinear function f has exponential critical growth, and potential V is a continuous function. By using the constrained variational methods, quantitative Deformation Lemma and Brouwer degree theory, we prove the existence of least energy sign-changing solutions.

    Citation: Kun Cheng, Wentao Huang, Li Wang. Least energy sign-changing solution for a fractional p-Laplacian problem with exponential critical growth[J]. AIMS Mathematics, 2022, 7(12): 20797-20822. doi: 10.3934/math.20221140

    Related Papers:

    [1] Qiang Li, Weiqiang Gong, Linzhong Zhang, Kai Wang . Robust dissipativity and passivity of stochastic Markovian switching CVNNs with partly unknown transition rates and probabilistic time-varying delay. AIMS Mathematics, 2022, 7(10): 19458-19480. doi: 10.3934/math.20221068
    [2] Sunisa Luemsai, Thongchai Botmart, Wajaree Weera, Suphachai Charoensin . Improved results on mixed passive and H performance for uncertain neural networks with mixed interval time-varying delays via feedback control. AIMS Mathematics, 2021, 6(3): 2653-2679. doi: 10.3934/math.2021161
    [3] Jenjira Thipcha, Presarin Tangsiridamrong, Thongchai Botmart, Boonyachat Meesuptong, M. Syed Ali, Pantiwa Srisilp, Kanit Mukdasai . Robust stability and passivity analysis for discrete-time neural networks with mixed time-varying delays via a new summation inequality. AIMS Mathematics, 2023, 8(2): 4973-5006. doi: 10.3934/math.2023249
    [4] Wenlong Xue, Zhenghong Jin, Yufeng Tian . Stability analysis of delayed neural networks via compound-parameter -based integral inequality. AIMS Mathematics, 2024, 9(7): 19345-19360. doi: 10.3934/math.2024942
    [5] Yanshou Dong, Junfang Zhao, Xu Miao, Ming Kang . Piecewise pseudo almost periodic solutions of interval general BAM neural networks with mixed time-varying delays and impulsive perturbations. AIMS Mathematics, 2023, 8(9): 21828-21855. doi: 10.3934/math.20231113
    [6] Huahai Qiu, Li Wan, Zhigang Zhou, Qunjiao Zhang, Qinghua Zhou . Global exponential periodicity of nonlinear neural networks with multiple time-varying delays. AIMS Mathematics, 2023, 8(5): 12472-12485. doi: 10.3934/math.2023626
    [7] Patarawadee Prasertsang, Thongchai Botmart . Improvement of finite-time stability for delayed neural networks via a new Lyapunov-Krasovskii functional. AIMS Mathematics, 2021, 6(1): 998-1023. doi: 10.3934/math.2021060
    [8] Rakkiet Srisuntorn, Wajaree Weera, Thongchai Botmart . Modified function projective synchronization of master-slave neural networks with mixed interval time-varying delays via intermittent feedback control. AIMS Mathematics, 2022, 7(10): 18632-18661. doi: 10.3934/math.20221025
    [9] Chantapish Zamart, Thongchai Botmart, Wajaree Weera, Prem Junsawang . Finite-time decentralized event-triggered feedback control for generalized neural networks with mixed interval time-varying delays and cyber-attacks. AIMS Mathematics, 2023, 8(9): 22274-22300. doi: 10.3934/math.20231136
    [10] Saravanan Shanmugam, R. Vadivel, Mohamed Rhaima, Hamza Ghoudi . Improved results on an extended dissipative analysis of neural networks with additive time-varying delays using auxiliary function-based integral inequalities. AIMS Mathematics, 2023, 8(9): 21221-21245. doi: 10.3934/math.20231082
  • In this paper, we consider the following fractional p-Laplacian equation involving Trudinger-Moser nonlinearity:

    (Δ)sN/su+V(x)|u|Ns2u=f(u)  in RN,

    where s(0,1),2<Ns=p. The nonlinear function f has exponential critical growth, and potential V is a continuous function. By using the constrained variational methods, quantitative Deformation Lemma and Brouwer degree theory, we prove the existence of least energy sign-changing solutions.



    Recently, neural networks (NNs) have drawn considerable attention in many fields of science and engineering applications for example associative memories, fixed-point computation, control, static image processing and combinatorial optimization Ref. [1,2,3]. However, time-delay is common in various biological and physical phenomena, which is demonstrated by applying of mathematical modelling with time-delay in a wide range of applications for instance mechanical transmission, fluid transmission, metallurgical processes and networked control systems which is frequently a source of chaos, instability and poor control performance. These applications are extensively dependent upon the stability of the equilibrium of NNs. So, stability is much importance in dynamical properties of NNs when NNs are designed. As research results, the stability problem and the performance of the NNs with time-delay have been improved in Ref. [4,5,6,7,8,9,10,11,12]. However, most results were discussed only on the discrete delay in NNs. In contrast, the distributed delay should be associated with a model of a system that there exists a distribution of propagation delays over a period of time in some cases as discussed in Ref. [13,14]. Therefore, there has been an increasing interest in the delayed NNs, and a great number of results on these topics have been reported in the literature Ref. [15,16,17,18,19,20,21,22,23] as well.

    On the other hand, the passivity is interesting problem and is closely related to the circuit analysis method. The properties of the passivity are that the system can keep the system internally stable Ref. [24,25]. Especially, the passive system employs the product of input and output as the energy provision and embodies the energy attenuation character. A passive system only burns energy without energy production and passivity represents the property of energy consumption Ref. [26]. The issue of passivity performance analysis has been used in various areas such as fuzzy control, signal processing, networked control and sliding mode control Ref. [27]. Due to these features, the passivity problems have been an active area of research in recently decades with NNs.

    In the same way, Ref. [28,29,30] also studied the passivity analysis of neural network with discrete and distributed delays. In addition, many uncertain factors such as uncertain parameters, disturbance and environmental noise are regularly encountered in many practical and engineering systems, and these make it difficult to develop an exact mathematical model. Therefore, the parameter uncertainties are very important and unavoidable while modelling NNs in both theoretical and practical cases. Meanwhile, improved the delay-dependent approach to passivity analysis for uncertain NNs with discrete interval and distributed time-varying delays has also discussed in Ref. [31,32]. However, there are few results for studying this problem with uncertainties to the best of the authors's knowledge, we study delay-dependent passivity criteria for uncertain NNs with discrete interval and distributed time-varying delays.

    Recently, several approaches to reduce the conservatism for the system with time delay have been reported in the literature, namely an appropriate Lyapunov-Krasovskii functional method by utilizing information of the neuron activation function and some techniques to evaluate the bounds on some cross-terms product arising in the analysis of the delay-dependent stability problem such as integral inequality, refined Jensen's inequality, and free weighting matrices approach etc. These approaches will give better maximum allowable upper bound for time-varying delay over some existing ones Ref. [33,34,35,36,37,38,39,40]. However, these previous works still study on delay-derivative-dependent stability criteria. Practically time delays can occur in an irregular fashion such as sometimes the time-varying delays are not differentiable.

    Therefore in this paper, we have followed robust passivity analysis of NNs with interval nondifferentiable and distributed time-varying delays to obtain a better maximum bound value and to relax the derivative condition of delay. Moreover, system is assumed that the parameter uncertainties are norm-bounded for checking the passivity of the addressed NNs in LMIs, which can be checked numerically using the effective LMI toolbox in MATLAB. This is the first time that we apply the methods to study the networks model to reduce the condition of delays being non-differentiable delays. Moreover, the system can be turned into the delayed NNs proposed in Ref. [36,39,40] which means that this work is more general than them. Furthermore, the main ideas of this work are given as follows:

    The challenge of this paper is studying the new result on robust passivity analysis of NNs with non-differentiable mixed time-varying delays which mean that this work can be used for various systems with fast time-varying delays compared with previous works considered on differentiable delay (˙r(t)μ).

    The new Lyapunov Krasovskii functional establishes more relationships among different vectors, avoids the extra conservatism arising from estimating the time-varying delays and utilizes more information about the upper and lower bounds of the time delays existing in the systems.

    The new sufficient conditions based on refined Jensen-based inequalities proposed in Ref. [41], are less conservative than the others proposed Ref. [33,34,35,36,37,38,39,40] which are shown in the comparison examples.

    The rest of paper is organized as follows: Section 2 provides some mathematical preliminaries and network model. Section 3 presents the passivity analysis of uncertain NNs with interval and distributed time-varying delays. Numerical examples are given in Section 4. Finally, the conclusion is provided in Section 5.

    Notations: Rn is the n-dimensional Euclidean space; Rm×n denotes the set of m×n real matrices; In represents the n-dimensional identity matrix. Let S+n denotes the set of symmetric positive definite matrices in Rn×n. We also denoted by D+n the set of positive diagonal matrices. A matrix D=diag{d1,d2,...,dn}D+n if di>0 (i=1,2,...,n). The notation X0 (respectively, X>0) means that X is positive semi-definite (respectively, positive definite); diag() denotes a block diagonal matrix; [XYZ] stands for [XYYTZ]; Matrix dimensions, if not explicitly stated, are assumed to be compatible for algebraic operations.

    Consider the following of NNs with nondifferentiable interval and distributed time-varying delays in the form:

    {˙p(t)=Dp(t)+Ag(p(t))+A1g(p(tr(t)))+A2ttd(t)g(p(s))ds+u(t),q(t)=C1g(p(t))+C2g(p(tr(t)))+C3ttd(t)g(p(s))ds+C4u(t),p(t)=ϕ(t),t[θ,0],θ=max{r2,d}, (2.1)

    where n denotes the number of neurons in the network, p(t)=[p1(t),p2(t),...,pn(t)]T Rn is the neurons state vector, q(t)Rn is the output vector and u(t) is the external input of the network, D=diag{d1,d2,...,dn} is a positive diagonal matrix, A,A1,A2 are interconnection weight matrices, C1,C2,C3,C4 are real matrices, g(p(t))=[g1(p1(t)),g2(p2(t)),...,gn(pn(t))]TRn denotes the activation function, g(p(tr(t)))=[g1(p1(tr(t))),g2(p2(tr(t))),...,gn(pn(tr(t)))]TRn. and ϕ(t)Rn is the initial function.

    The variables r(t) and d(t) represent the mixed delays of the model in (2.1) and satisfy

    0r1r(t)r2and0d(t)d,t0, (2.2)

    where r1,r2,andd are constants.

    The neural activation functions gi(pi(t)) are continuous gi(0)=0 and there exist constants li,l+i (i=1,2,...,n) such that

    ligi(p)gi(q)pql+i,p,qR, pq. (2.3)

    Definition 2.1. [5] The neural network (2.1) is said to be passive if there exists a scalar γ>0 such that for all tf0

    2tf0qT(s)u(s)ds  γtf0uT(s)u(s)ds, (2.4)

    under the zero initial condition.

    Lemma 2.2. [41] For a given matrix QS+n and a function e:[u,v]Rn whose derivative ˙eC([u,v],Rn), the following inequalities hold:

    vu˙eT(s)Q˙e(s)ds  1vuˆεTˉQ ˆε, (2.5)
    vuvs˙eT(α)Q˙e(α)dαds  2ˆΓTˆQ ˆΓ, (2.6)

    where ˉQ=diag{Q,3Q,5Q}, ˆQ=diag{Q,2Q}, ˆε=[εT1,εT2,εT3]T, ˆΓ=[ΓT1,ΓT2]T and

    ε1=e(v)e(u),ε2=e(v)+e(u)2vuvue(s)ds,ε3=e(v)e(u)+6vuvue(s)ds12(vu)2vuvse(α)dαds,Γ1=e(v)1vuvue(s)ds,Γ2=e(v)+2vuvue(s)ds6(vu)2vuvse(α)dαds.

    Lemma 2.3. [6] For a positive definite matrix P>0, and an integral function {e(α)|α[u,v]}, then the following inequalities hold:

    vueT(α)Pe(α)dα    1vu(vue(α)dαg)TP (vue(α)dαg), (2.7)
    vuvβvseT(α)Pe(α)dαdβds        6(vu)3(vuvβvse(α)dαdβdsg)TP (vuvβvse(α)dαdβdsg). (2.8)

    Lemma 2.4. [6] Let M,NandF(t) be real matrices of appropriate dimensions with F(t) satisfying FT(t)F(t)I. Then for any scalar ϵ>0.

    MF(t)N+(MF(t)N)T  ϵ1MMT+ϵNTN. (2.9)

    Lemma 2.5. [6] Given constant symmetric matrices P,Q,R with appropriate dimensions satisfying P=PT, Q=QT>0. Then P+RTQ1R<0 if and only if

    [PRTRQ]<0or[QRRTP]<0. (2.10)

    In this section, the new result on robust passivity analysis for NNs with interval nondifferentiable and distributed time-varying delays will be established. Let us set

    L1=diag{l1,l2,...,ln},     L2=diag{l+1,l+2,...,l+n},     r12=r2r1,

    ˉSk=diag{Sk,3Sk,5Sk}, k=1,2,     T=[InIn00InIn2In0InIn6In6In],

    Σ1(r)=[φT1r1φT9(rr1)φT10+(r2r)φT11r212φT12]T,

    Σ2=[AT(φ1φ2)T(φ2φ4)Tr1(φ1φ9)T]T,

    Σ3=[φT2φT8]T,    Σ4=[φT4φT7]T,    Σ5=[φT1φT5]T,

    Σ6=[φT1φT2φT9φT12]T,    Σ7=[φT3φT4φT11φT14]T,    Σ8=[φT2φT3φT10φT13]T,

    Σ9(r)=[((rr1)φ10+(r2r)φ11)T(φ16+φ17)T]T,

    Σ10=r212φ1r222φ12,     Σ11(r)=r2122φ2(rr1)22φ13(r2r)22φ14,

    Σ12=φ5L1φ1,Σ13=L2φ1φ5,

    Σ14=φ6L1φ3,Σ15=L2φ3φ6,

    Σ16=φ5φ6L1(φ1φ3),Σ17=L2(φ1φ3)φ5+φ6,

    Σ18=φ5φ8L1(φ1φ2),Σ19=L2(φ1φ2)φ5+φ8,

    Σ20=φ5φ7L1(φ1φ4),Σ21=L2(φ1φ4)φ5+φ7,

    Σ22=φ7φ8L1(φ4φ2),Σ23=L2(φ4φ2)φ7+φ8,

    Σ24=φ6φ8L1(φ3φ2),Σ25=L2(φ3φ2)φ6+φ8,

    Σ26=φ7φ6L1(φ4φ3),Σ27=L2(φ4φ3)φ7+φ6,

    Γ0(r)=ΣT1(r)PΣ2+φT1(L2W2L1W1)AφT5(W2W1)A,

    Γ1=φT1(Q2+Q3)φ1φT2Q2φ2φT4Q3φ4+ΣT3Q1Σ3ΣT4Q1Σ4+r212ΣT5S3Σ5+d2φT5S4φ5,

    Γ2=AT[r21S1+r212S2+r212R1+r2122R2+r6136Z1+(r32r31)(r2r1)336Z2]A,

    Γ3=φT15(γIn+2C4)φ15+Sym(φT15(C1φ5+C2φ6+C3φ18)),

    Ω1(r)=Sym(Γ0(r))+Γ1+Γ2Γ3,

    Ω2=ΣT6TTˉS1TΣ6,    Ω3=ΣT7TTˉS2TΣ7,    Ω4=ΣT8TTˉS2TΣ8,

    Ω5=2(φ1φ9)TR1(φ1φ9)+4(φ1+2φ93φ12)TR1(φ1+2φ93φ12),

    Ω6=2(φ2φ10)TR2(φ2φ10)+4(φ2+2φ103φ13)TR2(φ2+2φ103φ13)+2(φ3φ11)TR2(φ3φ11)+

    4(φ3+2φ113φ14)TR2(φ3+2φ113φ14),

    Π(r)=ΣT9(r)S3Σ9(r)+ΣT10Z1Σ10+ΣT11(r)Z2Σ11(r),

    Δ0=ΣT12Δ1Σ13+ΣT14Δ2Σ15+ΣT16Δ3Σ17+ΣT18Δ4Σ19+ΣT20Δ5Σ21+ΣT22Δ6Σ23+ΣT24Δ7Σ25+ΣT26Δ8Σ27,

    A=Dφ1+Aφ5+A1φ6+A2φ18+φ15,

    and φi=[0n×(i1)nIn0n×(18i)n](i=1,2,...,18).

    Based on the Lyapunov–Krasovskii functional approach, we present our new theorem for passivity of NNs (2.1).

    Theorem 3.1. The delayed neural network in (2.1) is passive in the sense of definition 2.1 for any delays r(t)andd(t) satisfying 0r1r(t)r2 and 0d(t)d if there exist matrices PS+4n;Q1,S3S+2n;Q2,Q3,S1,S2,S4,R1,R2,Z1,Z2S+n;Δk,WσD+n, (k=1,2,...,8;σ=1,2), and a scalar γ>0 satisfy the following LMI:

    Φ(r) = Sym(Δ0)Π(r)Ω1(r)6k=2Ωk < 0. (3.1)

    Proof. Consider the following Lyapunov-Krasovskii functional:

    V(t,pt)=5i=1Vi(t,pt),

    where,

    V1(t,pt)=ηT1(t)Pη1(t)+2ni=1w1ipi(t)0(gi(s)lis)ds+2ni=1w2ipi(t)0(l+isgi(s))ds,V2(t,pt)=tr1tr2ηT2(s)Q1η2(s)ds+ttr1pT(s)Q2p(s)ds+ttr2pT(s)Q3p(s)ds,V3(t,pt)=r10r1tt+s˙pT(u)S1˙p(u)duds+r12r1r2tt+s˙pT(u)S2˙p(u)duds+r12r1r2tt+sηT2(u)S3η2(u)duds+d0dtt+sgT(p(u))S4g(p(u))duds,V4(t,pt)=ttr1tstu˙pT(λ)R1˙p(λ)dλduds+r1r2r1stt+u˙pT(λ)R2˙p(λ)dλduds,V5(t,pt)=r316ttr1tstλtu˙pT(θ)Z1˙p(θ)dθdudλds+(r32r31)6r1r2r1sr1λtt+u˙pT(θ)Z2˙p(θ)dθdudλds.

    Let Δk=diag{λk1,λk2,...,λkn}(k=1,2,...,8), Wσ=diag{wσ1,wσ2,...,wσn} (j=1,2), and

    η1(t)=[pT(t)ttr1pT(s)dstr1tr2pT(s)dsttr1tspT(u)duds]T,η2(t)=[pT(t)gT(p(t))]T,ξ(t)=[pT(t),pT(tr1),pT(tr(t)),pT(tr2),gT(p(t)),gT(p(tr(t))),gT(p(tr2)),gT(p(tr1)),1r1ttr1pT(s)ds,1r(t)r1tr1tr(t)pT(s)ds,1r2r(t)tr(t)tr2pT(s)ds,2r21ttr1tspT(u)duds,2(r(t)r1)2tr1tr(t)tr1spT(u)duds,2(r2r(t))2tr(t)tr2tr(t)spT(u)duds,uT(t),tr1tr(t)gT(p(s))ds,tr(t)tr2gT(p(s))ds,ttd(t)gT(p(s))ds]T.

    The derivative of V(t,pt) along the solution of system (2.1) as follows:

    ˙V1(t,pt)=2ηT1(t)P˙η1(t)+2ni=1w1i[gi(pi(t))˙pi(t)li(pi(t))˙pi(t)]+2ni=1w2i[l+i(pi(t))˙pi(t)gi(pi(t))˙pi(t)]=2ηT1(t)P˙η1(t)+2[pT(t)(L2W2L1W1)gT(p(t))(W2W1)]˙p(t),˙V2(t,pt)=ηT2(tr1)Q1η2(tr1)ηT2(tr2)Q1η2(tr2)+pT(t)(Q2+Q3)p(t)pT(tr1)Q2p(tr1)pT(tr2)Q3p(tr2),˙V3(t,pt)=r21˙pT(t)S1˙p(t)+r212˙pT(t)S2˙p(t)+r212ηT2(t)S3η2(t)r1ttr1˙pT(s)S1˙p(s)dsr12tr1tr2˙pT(s)S2˙p(s)ds+d2gT(p(t))S4g(p(t))r12tr1tr2ηT2(s)S3η2(s)dsdttdgT(p(s))S4g(p(s))ds,˙V4(t,pt)=r212˙pT(t)R1˙p(t)+r2122˙pT(t)R2˙p(t)ttr1ts˙pT(u)R1˙p(u)dudsr1r2tr1t+s˙pT(u)R2˙p(u)duds,˙V5(t,pt)=r6136˙pT(t)Z1˙p(t)+(r32r31)(r2r1)336˙pT(t)Z2˙p(t)r316ttr1tstλ˙pT(u)Z1˙p(u)dudλds(r32r31)6r1r2r1str1t+λ˙pT(u)Z2˙p(u)dudλds.

    We conclude that,

    ˙V(t,pt)ξT(t)(Sym(Γ0(r))+Γ1+Γ2)ξ(t)r1ttr1˙pT(s)S1˙p(s)dsr12tr1tr2˙pT(s)S2˙p(s)dsr12tr1tr2ηT2(s)S3η2(s)dsd(t)ttd(t)gT(p(s))S4g(p(s))dsttr1ts˙pT(u)R1˙p(u)dudstr1tr2tr1s˙pT(u)R2˙p(u)dudsr316ttr1tstλ˙pT(u)Z1˙p(u)dudλds(r32r31)6r1r2r1str1t+λ˙pT(u)Z2˙p(u)dudλds. (3.2)

    According to Lemma 2.2, we have

    r1ttr1˙pT(s)S1˙p(s)dsξT(t)ΣT6TT¯S1TΣ6ξ(t)=ξT(t)Ω2ξ(t). (3.3)

    By splitting, we have

    r12tr1tr2˙pT(s)S2˙p(s)ds=r12tr(t)tr2˙pT(s)S2˙p(s)dsr12tr1tr(t)˙pT(s)S2˙p(s)ds. (3.4)

    Applying Lemma 2.2 yields

    r12tr(t)tr2˙pT(s)S2˙p(s)ds   ξT(t)ΣT7TT¯S2TΣ7ξ(t)=ξT(t)Ω3ξ(t). (3.5)
    r12tr1tr(t)˙pT(s)S2˙p(s)ds   ξT(t)ΣT8TTˉS2TΣ8ξ(t)=ξT(t)Ω4ξ(t). (3.6)
    ttr1ts˙pT(u)R1˙p(u)dudsξT(t)Ω5ξ(t). (3.7)
    tr1tr2tr1s˙pT(u)R2˙p(u)dudsξT(t)Ω6ξ(t). (3.8)

    In the same way, applying the inequalities (2.7) and (2.8), then we obtain

    r12tr1tr2ηT2(s)S3η2(s)ds(tr1tr2η2(s)ds)TS3(tr1tr2η2(s)ds)ξT(t)ΣT9(r)S3Σ9(r)ξ(t), (3.9)
    d(t)ttd(t)g(p(s))T(s) S4 g(p(s))ds(ttd(t)g(p(s))ds)TS4(ttd(t)g(p(s))ds)=ξT(t)φT18S4φ18ξ(t), (3.10)
    r316ttr1tstλ˙pT(u)Z1˙p(u)dudλdsξT(t)(r212φ1r212φ12)TZ1(r212φ1r212φ12)ξ(t)=ξT(t)ΣT10Z1Σ10ξ(t), (3.11)
    (r32r31)6r1r2r1str1t+λ˙pT(u)Z2˙p(u)dudλdsξT(t)ΣT11(r)Z2Σ11(r)ξ(t). (3.12)

    For λ1i>0,i=1,2,,n, it can be deduced from (2.3) that

    2(gi(pi(t))lipi(t))λ1i(l+ipi(t)gi(pi(t)))0,

    and thus

    ξT(t)Sym(ΣT12Δ1Σ13)ξ(t)0. (3.13)

    From (3.13), we have

    ξT(t)Sym(Δ0)ξ(t)0. (3.14)

    Then, to show that NNs (2.1) is passive, we define J(tf)=tf0[γuT(t)u(t)2qT(t)u(t)]dt where tf0. Consider the zero initial condition and we have

    J(tf)=tf0[˙V(pt)γuT(t)u(t)2qT(t)u(t)]dtV(ptf)tf0[˙V(pt)γuT(t)u(t)2qT(t)u(t)dt].

    From (3.2) to (3.14), it can be deduced that

    ˙V(pt)γuT(t)u(t)2qT(t)u(t)ξT(t)Φ(r)ξ(t).

    where Φ(r) is an affine function in r, for Φ(r)<0, r[r1,r2] if and only if Φ(r1)<0 and Φ(r2)<0. if (3.1) holds for r=r1 and r=r2 and we have Φ(r)<0, then

    ˙V(pt)γuT(t)u(t)2qT(t)u(t)  0.

    Considering, we have J(tf)<0 for any tf0 if condition (2.3) is satisfied. Thus, the system of NNs (2.1) is passive. The proof is completed.

    Remark 1. We can see that the time delay in this work is a continuous function which belongs to a given interval. It means that the lower and upper bounds of the time-varying delay are available. Moreover, there is no need to be differentiable for the delay function. Therefore, the delays considering in this brief are more general than those studied in [29,33,34,38].

    Remark 2. The activation function in inequality (2.3) studied by [40] is more general than [28,33,36,39] because the constants li and l+i can be positive, zero or negative. We can see that the activation function under (2.3) can be unbounded, non-monotonic, non-differentiable. Hence, the passivity condition is considered in this work is less conservative than Ref. [28,33,36,39].

    Remark 3. The proof of theorem 3.1 shows estimating of integral terms by lemma 2.2 applying equations (3.3), (3.4), (3.5), (3.6), (3.7) and (3.8), which obtained a tighter upper bound than Jensen's inequality used in Ref. [25,29,33,36].

    Based on the presented passivity condition in theorem 3.1, we will develop passivity analysis of uncertain NNs established as follows. Consider

    {˙p(t)=(D+ΔD(t))p(t)+(A+ΔA(t))g(p(t))+(A1+ΔA1(t))g(p(tr(t)))             +(A2+ΔA2(t))ttd(t)g(p(s))ds+u(t),q(t)=C1g(p(t))+C2g(p(tr(t)))+C3ttd(t)g(p(s))ds+u(t),p(t)=ϕ(t),  t[θ,0],θ=max{r2,d}, (3.15)

    where ΔD(t),ΔA(t),ΔA1(t),ΔA2(t) are the time-varying parameter uncertainties, which are assumed to be of the form

    [ΔD(t)ΔA(t)ΔA1(t)ΔA2(t)]=MF(t)[N1N2N3N4], (3.16)

    where M,N1,N2,N3andN4 are known real constant matrices, and F() is an unknown time-varying matrix function satisfying FT(t)F(t)I then we have the following result.

    Theorem 3.2. The delayed neural network in (3.15) is passive in the sense of definition 2.1 for any delays τ(t)andd(t) satisfying 0r1r(t)r2 and 0d(t)d if there exist matrices PS+4n;Q1,S3S+2n;Q2,Q3,S1,S2,S4,R1,R2,Z1,Z2S+n;Δk,WσD+n, (k=1,2,......,8;σ=1,2), and a scalar γ>0 satisfy the following LMI:

    [Φ(r)+ϵΞT2Ξ2ΞT1Ξ1ϵI]<0, (3.17)

    where

    Ξ1=[P1M000000000000000000],Ξ2=[N1000N2N300000000000N40],

    and Φ(r) is defined in theorem 3.1.

    Proof. Replacing D,A,A1,andA2 in theorem 3.1 with D+MF(t)N1, A+MF(t)N2,A1+MF(t)N3,A2+MF(t)N4 respectively, so we have

    Φ(r)+ΞT1F(t)Ξ2+ΞT2F(t)Ξ1<0. (3.18)

    Applying lemma 2.4, it can be deduced that for ϵ>0

    Φ(r)+ϵ1ΞT1Ξ1+ϵΞT2Ξ2<0. (3.19)

    From lemma 2.5 shows that (3.19) is equivalent to (3.17), therefore the proof is completed.

    Remark 4. If delayed NNs (2.1) are setting as C2=0,C3=0 and C4=0, the networks model turns into the delayed NNs proposed in [28,33,38,40]:

    {˙p(t)=Dp(t)+Ag(p(t))+A1g(p(tr(t)))+A2ttd(t)g(p(s)) ds+u(t),q(t)=C1g(p(t)),p(t)=ϕ(t),  t[d,0]. (3.20)

    Hence, our network model (2.1) includes previous network model, which can be regarded as a special case of neural network (2.1).

    Remark 5. To illuminate how to solve the upper bound of r2 for system (2.1) satisfying time-varying delays (2.2) and neural activation functions (2.3), the following steps are performed.

    Step 1: Given positive diagonal matrix D, real matrices A,A1,A2,C1,C2,C3 and positive constants r1,d.

    Step 2: Select a positive constant γ.

    Step 3: Define variable matrices with appropriate dimensions P,Q1,S3,Q2,Q3,S1,S2,S4,R1,R2, Z1,Z2,Δk,Wσ (k=1,2,...,8;σ=1,2).

    Step 4: Use matlab software to compute the value of the variable.

    Step 5: Calculate the value of LMIs, in (3.1).

    In this section, three numerical examples are given to illustrate the merits of the proposed robust passivity results.

    Example 4.1. Consider a neural network (2.1) with the following parameters:

    D=[2.2001.8], A=[1.210.20.3], A1=[0.80.40.20.1], A2=[0000],

    L1=diag{0,0}, L2=diag{1,1}, C1=I, and C2=C3=C4=0. The neural activation functions are assumed to be gi(pi)=12(|pi+1||pi1|) (i=1,2). It is easy to check that the neural activation functions are satisfied (2.3) with li=0 and l+i=0 (i=1,2). Using Matlab LMI Toolbox, we can conclude that the upper bound of r2 without non differentiable μ which is shown in Table 1 is feasibility of the LMI in theorem 3.1. In addition, the results from [36,37,38,39,40] without distributed delay are listed in Table 1. As shown in this table, the criterion of this paper is less conservative than those results obtained in [36,37,38,39,40]. According to Figure 1, it can be confirmed that neural network (2.1) under zero input and the initial condition [p1(t),p2(t)]T=[1,1]T is stable.

    Table 1.  Upper bound of r2 for Example 4.1.
    μ μ=0.5 unknown
    [36] 0.5227 -
    [39] 1.3752 -
    [40] 3.0430 -
    [38] 3.0835 -
    [37] 3.6566 -
    Theorem 3.1 - 4.1010

     | Show Table
    DownLoad: CSV
    Figure 1.  State trajectory of neural network in Example 4.1.

    Example 4.2. Consider a uncertain neural network (3.15) with the following parameters:

    D=[2.3002.5], A=[0.30.20.40.1], A1=[0.50.70.70.4], A2=[0.50.30.21.2],

    M=diag{0.1,0.1}, N1=N2=N3=N4=diag{1,1}. With these parameters, we can conclude that the upper bound of r2 are shown in Table 2 is feasibility of the LMI in theorem 3.2. Moreover, the results from [33,35,40] are listed in Table 2. As shown in this table, the criteria of this paper is less conservative than those results obtained in [33,35,40]. We have activation functions as above and set ΔD(t)=ΔA(t)=ΔA1(t)=ΔA2(t)=[0.1sin(t)000.1sin(t)] shown in Figure 2. From Figure 2, it can be confirmed that the neural network (3.15) without input u(t) is robustly stable, which the initial condition [p1(t),p2(t)]T=[1.5,1.5]T.

    Table 2.  Upper bound of r2 for Example 4.2.
    μ=0.1 unknown μ
    [33] 0.5005 0.4269
    [40] 0.5504 -
    [35] 0.6621 -
    Theorem 3.2 - 3.0420

     | Show Table
    DownLoad: CSV
    Figure 2.  State trajectory of neural network in Example 4.2.

    Example 4.3. Consider a uncertain neural network (3.15) with the following parameters:

    D=[2.2001.5], A=[10.60.10.3], A1=[10.10.10.2], A2=[0000],

    M=diag{0.1,0.1}, N1=0.1,N2=0.2I,N3=0.3I,N4=diag{0,0} and C1=C2=C3=C4=I. In this example, we can conclude that the upper bounds of r2 are shown in Table 3 is feasibility of the LMI in theorem 3.2. Moreover, the results from [33,34,40] without distributed delay are listed in Table 3. As shown in this table, the criterion of this paper is less conservative than those results obtained in [33,34,40]. We have activation functions as above and set ΔD(t)=[0.01sin(t)000.01sin(t)],ΔA(t)=[0.02sin(t)000.02sin(t)],ΔA1(t)=[0.03sin(t)000.03sin(t)] shown in Figure 3. From Figure 3, it can be confirmed that the neural network (3.15) without input u(t) is robustly stable, which the initial condition [p1(t),p2(t)]T=[2,1]T.

    Table 3.  Upper bound of r2 for Example 4.3.
    μ=0.3 unknown μ
    [33] 0.4197 0.3994
    [40] 1.9091 -
    [34] 2.1350 -
    Theorem 3.2 - 2.3220

     | Show Table
    DownLoad: CSV
    Figure 3.  State trajectory of neural network in Example 4.3.

    Remark 6. An important property in linear circuit and system theory is passivity which is applicable to the analysis of properties of immittance or hybrid matrices of various classes of neural networks, inverse problem of linear optimal control, Popov criterion, circle criterion and spectral factorization by algebra [42]. In the recent years, passivity properties have also been related to the neural networks [36,37,38,39,40]. It should be pointed out that the aforementioned results have the restrictions on the derivative time-varying delays which mean that the delayed conditions in this work are more applicable in the real-world system by establishing Lyapunov-Krasovskii functional fully of the information of the delays r1,r2 and d. On the other hand, in this work, we use the refined Jensen's inequality to estimate single and double integrals. By applying the aforementioned techniques, we obtain the less conservative results than the others [25,29,33,36].

    Example 4.4. Consider a uncertain neural network (3.15) with the following parameters:

    D=[2.00002.50002.3],        A=[2.70.60.340.51.00.160.82.01.0], A1=[0.80.150.160.50.10.250.50.251.0],  A2=[0.370.91.380.180.10.130.50.70.5],

    M=diag{0.1,0.1,0.1}, N1=0.1I, N2=0.2I, N3=0.2I, N4=0.1I, C1=C2=C3=C4=I and

    ΔD(t)=[0.2sin(t)0000.2sin(t)0000.3sin(t)],ΔA(t)=[0.15sin(t)0000.05sin(t)0000.1sin(t)],
    ΔA1(t)=[0.15sin(t)0000.1sin(t)0000.05sin(t)],ΔA2(t)=[0.1sin(t)0000.2sin(t)0000.1sin(t)].

    In Example 4.4, the state trajectory of neural network (3.15) for r1=1,r2=2,d=0.5 and g1(s)=g2(s)=g3(s)=tanh(s) without input u(t) has been analyzed. The result is robustly stable shown in Figure 4 with the initial condition [p1(t),p2(t),p3(t)]T=[2,1,2,]T.

    Figure 4.  State trajectory of neural network in Example 4.4.

    In this research, we focused on new results for robust passivity analysis of NNs with interval nondifferentiable and distributed time-varying delays. Using refined Jensen's inequalities, and applying the Lyapunov-Krasovskii functional containing single, double, triple and quadruple integrals, the new conditions were obtained in terms of LMI which can be checked by using LMI toolbox in MATLAB. Moreover, These results are less conservative than the existing ones and can be an effective method. Compared with existing ones, the obtained criteria are more effective because of the application of refined Jensen-based inequality technique comprising single and double inequalities evaluating. Three numerical examples have been proposed to show the effectiveness of the methods. For further research, we can use these methods to consider the dynamic networks with Markovian jumping delayed complex networks or stochastic delayed complex networks.

    The first author was financially supported by the Thailand Research Fund (TRF), the Office of the Higher Education Commission (OHEC) (grant number : MRG6280149) and Khon Kaen University. The fourth author was financial supported by University of Pha Yao. The fifth author was supported by Rajamangala University of Technology Isan (RMUTI) and Thailand Science Research and Innovation (TSRI). Contract No.Contract No. FRB630010/0174-P6-03.

    The authors declare that there is no conflict of interests regarding the publication of this paper.



    [1] V. Ambrosio, Multiplicity of positive solutions for a class of fractional Schrödinger equations via penalization method, Ann. Mat. Pura Appl., 196 (2017), 2043–2062. https://doi.org/10.1007/s10231-017-0652-5 doi: 10.1007/s10231-017-0652-5
    [2] V. Ambrosio, T. Isernia, Multiplicity and concentration results for some nonlinear Schrödinger equations with the fractional p-Laplacian, Discrete Cont. Dyn., 38 (2018), 5835–5881. https://doi.org/10.3934/dcds.2018254 doi: 10.3934/dcds.2018254
    [3] V. Ambrosio, T. Isernia, Sign-changing solutions for a class of schrödinger equations with vanishing potentials, Atti Accad. Naz. Lincei Cl. Sci. Fis. Mat. Natur., 29 (2018), 127–152. https://doi.org/10.4171/RLM/797 doi: 10.4171/RLM/797
    [4] S. Barile, G. Figueiredo, Existence of least energy positive, negative and nodal solutions for a class of p&q-problems with potentials vanishing at infinity, J. Math. Anal. Appl., 427 (2015), 1205–1233. https://doi.org/10.1016/j.jmaa.2015.02.086 doi: 10.1016/j.jmaa.2015.02.086
    [5] T. Bartsch, Z. Liu, T. Weth, Sign changing solutions of superlinear Schrödinger equations, Commun. Part. Diff. Eq., 29 (2005), 25–42. https://doi.org/10.1081/PDE-120028842 doi: 10.1081/PDE-120028842
    [6] T. Bartsch, T. Weth, Three nodal solutions of singularly perturbed elliptic equations on domains without topology, Ann. Inst. H. Poincaré Anal. Non Linéaire, 22 (2005), 259–281. https://doi.org/10.1016/j.anihpc.2004.07.005 doi: 10.1016/j.anihpc.2004.07.005
    [7] T. Bartsch, T. Weth, M. Willem, Partial symmetry of least energy nodal solutions to some variational problems, J. Anal. Math., 96 (2005), 1–18. https://doi.org/10.1007/BF02787822 doi: 10.1007/BF02787822
    [8] E. Böer, O. Miyagaki, Existence and multiplicity of solutions for the fractional p-Laplacian Choquard logarithmic equation involving a nonlinearity with exponential critical and subcritical growth, J. Math. Phys., 62 (2021), 051507. https://doi.org/10.1063/5.0041474 doi: 10.1063/5.0041474
    [9] X. Cabré, Y. Sire, Nonlinear equations for fractional Laplacians, Ⅰ: regularity, maximum principles, and Hamiltonian estimates, Ann. Inst. H. Poincaré Anal. Non Linéaire, 31 (2014), 23–53. https://doi.org/10.1016/j.anihpc.2013.02.001 doi: 10.1016/j.anihpc.2013.02.001
    [10] L. Caffarelli, L. Silvestre, An extension problem related to the fractional laplacian, Commun. Part. Diff. Eq., 32 (2007), 1245–1260. https://doi.org/10.1080/03605300600987306 doi: 10.1080/03605300600987306
    [11] D. Cao, Nontrivial solution of semilinear elliptic equation with critical exponent in R2, Commun. Part. Diff. Eq., 17 (1992), 407–435. https://doi.org/10.1080/03605309208820848 doi: 10.1080/03605309208820848
    [12] A. Castro, J. Cossio, J. Neuberger, A sign-changing solution for a superlinear Dirichlet problem, Rocky Mountain J. Math., 27 (1997), 1041–1053. https://doi.org/10.1216/rmjm/1181071858 doi: 10.1216/rmjm/1181071858
    [13] X. Chang, Z. Nie, Z. Wang, Sign-changing solutions of fractional p-laplacian problems, Adv. Nonlinear Stud., 19 (2019), 29–53. https://doi.org/10.1515/ans-2018-2032 doi: 10.1515/ans-2018-2032
    [14] X. Chang, Z. Wang, Nodal and multiple solutions of nonlinear problems involving the fractional Laplacian, J. Differ. Equations, 256 (2014), 2965–2992. https://doi.org/10.1016/j.jde.2014.01.027 doi: 10.1016/j.jde.2014.01.027
    [15] M. de Souza, On a class of nonhomogeneous fractional quasilinear equations in RN with exponential growth, Nonlinear Differ. Equ. Appl., 22 (2015), 499–511. https://doi.org/10.1007/s00030-014-0293-y doi: 10.1007/s00030-014-0293-y
    [16] M. de Souza, U. Severo, T. do Rêgo, Nodal solutions for fractional elliptic equations involving exponential critical growth, Math. Method. Appl. Sci., 43 (2020), 3650–3672. https://doi.org/10.1002/mma.6145 doi: 10.1002/mma.6145
    [17] A. Di Castro, T. Kuusi, G. Palatucci, Local behavior of fractional p-minimizers, Ann. Inst. H. Poincaré Anal. Non Linéaire, 33 (2016), 1279–1299. https://doi.org/10.1016/j.anihpc.2015.04.003 doi: 10.1016/j.anihpc.2015.04.003
    [18] E. Di Nezza, G. Palatucci, E. Valdinoci, Hitchhiker's guide to the fractional Sobolev spaces, Bull. Sci. Math., 136 (2012), 521–573. https://doi.org/10.1016/j.bulsci.2011.12.004 doi: 10.1016/j.bulsci.2011.12.004
    [19] P. Felmer, A. Quaas, J. Tan, Positive solutions of the nonlinear Schrödinger equation with the fractional Laplacian, P. Roy. Soc. Edinb. A, 142 (2012), 1237–1262. https://doi.org/10.1017/S0308210511000746 doi: 10.1017/S0308210511000746
    [20] G. Franzina, G. Palatucci, Fractional p-eigenvalues, Riv. Mat. Uni. Parma, 5 (2014), 373–386.
    [21] A. Iannizzotto, S. Liu, K. Perera, M. Squassina, Existence results for fractional p-Laplacian problems via Morse theory, Adv. Calc. Var., 9 (2016), 101–125. https://doi.org/10.1515/acv-2014-0024 doi: 10.1515/acv-2014-0024
    [22] A. Iannizzotto, M. Squassina, 1/2-Laplacian problems with exponential nonlinearity, J. Math. Anal. Appl., 414 (2014), 372–385. https://doi.org/10.1016/j.jmaa.2013.12.059 doi: 10.1016/j.jmaa.2013.12.059
    [23] H. Kozono, T. Sato, H. Wadade, Upper bound of the best constant of a Trudinger-Moser inequality and its application to a Gagliardo-Nirenberg inequality, Indiana Univ. Math. J., 55 (2006), 1951–1974. https://doi.org/10.1512/iumj.2006.55.2743 doi: 10.1512/iumj.2006.55.2743
    [24] I. Kuzin, S. Pohozaev, Entire solutions of semilinear elliptic equations, Basel: Birkhäuser, 1997. https://doi.org/10.1007/978-3-0348-9250-6
    [25] N. Laskin, Fractional quantum mechanics and Lévy path integrals, Phys. Lett. A, 268 (2000), 298–305. https://doi.org/10.1016/S0375-9601(00)00201-2 doi: 10.1016/S0375-9601(00)00201-2
    [26] N. Laskin, Fractional Schrödinger equation, Phys. Rev. E, 66 (2002), 056108. https://doi.org/10.1103/PhysRevE.66.056108 doi: 10.1103/PhysRevE.66.056108
    [27] Q. Li, Z. Yang, Multiple solutions for a class of fractional quasi-linear equations with critical exponential growth in RN, Complex Var. Elliptic, 61 (2016), 969–983. https://doi.org/10.1080/17476933.2015.1131683 doi: 10.1080/17476933.2015.1131683
    [28] C. Miranda, Un'osservazione su un teorema di Brouwer, Boll. Un. Mat. Ital., 3 (1940), 5–7.
    [29] J. Moser, A sharp form of an inequality by N. Trudinger, Indiana Univ. Math. J., 20 (1970), 1077–1092.
    [30] T. Ozawa, On critical cases of Sobolev's inequalities, J. Funct. Anal., 127 (1995), 259–269. https://doi.org/10.1006/jfan.1995.1012 doi: 10.1006/jfan.1995.1012
    [31] R. Pei, Fractional p-Laplacian equations with subcritical and critical exponential growth without the Ambrosetti-Rabinowitz condition, Mediterr. J. Math., 15 (2018), 66. https://doi.org/10.1007/s00009-018-1115-y doi: 10.1007/s00009-018-1115-y
    [32] P. Pucci, M. Xiang, B. Zhang, Multiple solutions for nonhomogeneous Schrödinger-Kirchhoff type equations involving the fractional p-Laplacian in RN, Calc. Var., 54 (2015), 2785–2806. https://doi.org/10.1007/s00526-015-0883-5 doi: 10.1007/s00526-015-0883-5
    [33] S. Secchi, Ground state solutions for nonlinear fractional Schrödinger equations in RN, J. Math. Phys., 54 (2013), 031501. https://doi.org/10.1063/1.4793990 doi: 10.1063/1.4793990
    [34] N. Thin, Multiplicity and concentration of solutions to a fractional p-Laplace problem with exponential growth, J. Math. Anal. Appl., 506 (2022), 125667. https://doi.org/10.1016/j.jmaa.2021.125667 doi: 10.1016/j.jmaa.2021.125667
    [35] Z. Wang, H. Zhou, Radial sign-changing solution for fractional Schrödinger equation, Discrete Cont. Dyn., 36 (2016), 499–508. https://doi.org/10.3934/dcds.2016.36.499 doi: 10.3934/dcds.2016.36.499
    [36] M. Willem, Minimax theorems, Boston: Birkhäuser, 1996. https://doi.org/10.1007/978-1-4612-4146-1
    [37] C. Zhang, Trudinger-Moser inequalities in fractional Sobolev-Slobodeckij spaces and multiplicity of weak solutions to the fractional-Laplacian equation, Adv. Nonlinear Stud., 19 (2019), 197–217. https://doi.org/10.1515/ans-2018-2026 doi: 10.1515/ans-2018-2026
  • This article has been cited by:

    1. Thanasak Mouktonglang, Kanyuta Poochinapan, Suriyon Yimnet, Robust Finite-Time Control of Discrete-Time Switched Positive Time-Varying Delay Systems with Exogenous Disturbance and Their Application, 2022, 14, 2073-8994, 735, 10.3390/sym14040735
    2. Thanasak Mouktonglang, Suriyon Yimnet, Muhammad Gulzar, Global Exponential Stability of Both Continuous-Time and Discrete-Time Switched Positive Time-Varying Delay Systems with Interval Uncertainties and All Unstable Subsystems, 2022, 2022, 2314-8888, 1, 10.1155/2022/3968850
    3. Li-Juan Liu, Cheng-En Zhou, Yong Ning, Jing-Chao Xu, Yucheng Jiang, 2024, Robust stability analysis of uncertain switched systems with linear time-varying delays, 9781510681651, 11, 10.1117/12.3037401
    4. Thanasak Mouktonglang, Kanyuta Poochinapan, Suriyon Yimnet, A Switching Strategy for Stabilization of Discrete-Time Switched Positive Time-Varying Delay Systems with All Modes Being Unstable and Application to Uncertain Data, 2023, 12, 2075-1680, 440, 10.3390/axioms12050440
    5. Nayika Samorn, Kanit Mukdasai, Issaraporn Khonchaiyaphum, Analysis of finite-time stability in genetic regulatory networks with interval time-varying delays and leakage delay effects, 2024, 9, 2473-6988, 25028, 10.3934/math.20241220
    6. Jiayang Liu, Yantao Wang, Xin Wang, Passivity analysis of neutral‐type Cohen–Grossberg neural networks involving three kinds of time‐varying delays, 2024, 0170-4214, 10.1002/mma.10332
  • Reader Comments
  • © 2022 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Metrics

Article views(1740) PDF downloads(57) Cited by(0)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog